The measurement of gastric intramucosal pH serves as a non-invasive technique for early detection of gastrointestinal ischaemia in critically ill patients. The method is based on the determination of the partial pressure of carbon dioxide in a 0.9% saline solution using a standard blood gas analyser. However, the use of standard blood gas analysers leads to an underestimation of carbon dioxide partial pressure in saline. Instrumental biases of six blood gas analysers were investigated using either a saline or a phosphate-buffered solution. Both test solutions were equilibrated with five defined carbon dioxide concentrations. Each blood gas analyser underestimated this defined partial pressure of carbon dioxide with a bias between -3.7% and -57.5% if saline was used. The phosphate-buffered solution considerably improved instrumental precision, resulting in biases between +2.7% and -17.6% Thus, a phosphate-buffered solution increases the accuracy of gastric intramucosal pH measurement.